Improved Transferability of Data‐Driven Damage Models Through Sample Selection Bias Correction

Damage models for natural hazards are used for decision making on reducing and transferring risk. The damage estimates from these models depend on many variables and their complex sometimes nonlinear relationships with the damage. In recent years, data‐driven modeling techniques have been used to capture those relationships. The available data to build such models are often limited. Therefore, in practice it is usually necessary to transfer models to a different context. In this article, we show that this implies the samples used to build the model are often not fully representative for the situation where they need to be applied on, which leads to a “sample selection bias.” In this article, we enhance data‐driven damage models by applying methods, not previously applied to damage modeling, to correct for this bias before the machine learning (ML) models are trained. We demonstrate this with case studies on flooding in Europe, and typhoon wind damage in the Philippines. Two sample selection bias correction methods from the ML literature are applied and one of these methods is also adjusted to our problem. These three methods are combined with stochastic generation of synthetic damage data. We demonstrate that for both case studies, the sample selection bias correction techniques reduce model errors, especially for the mean bias error this reduction can be larger than 30%. The novel combination with stochastic data generation seems to enhance these techniques. This shows that sample selection bias correction methods are beneficial for damage model transfer.     

People with disability and new disaster communications: access and the social media mash-up

This article explores how a lack of access to increasingly complex and overlapping digital communications platforms in times of disaster for people with disabilities has the potential to make already life-threatening situations considerably more dangerous. As we are increasingly coming to rely on a social media mash-up of digital platforms to assist in communications during disaster situations, the issue of accessibility for people with disabilities is as dire as if it was high ground during a tsunami or transport during a typhoon. The contemporary social media environment is characterised by a complex and overlapping network of complementary platforms, populated by user-generated content, where people communicate and exchange ideas. In this environment, YouTube videos are posted to Facebook and embedded in blogs, and Twitter is used to link to these other sites and is itself embedded in other platforms. These networks are increasingly supplementing and supplanting more traditional communication platforms, such as the television and radio, particularly in times of disaster. The concern of this paper is that the elements from which this mash-up of communications channels is made are not always accessible to people with disabilities. This evolving network of social media-based communication exposes the limits of existing Internet-based universal design.     

构建立足“3R”的矿难治理预防管理体系

对劳动力载体的保护是劳动和社会保障的指向之一。面对矿难对劳动者生存权的侵蚀,加强矿难治理的准备工作,就成了生命对社会的合理诉求。这其中对劳动者生存权的尊重,也是构建和谐社会的应有之义。     

日本都道府县防灾无线网的建设及借鉴

日本是世界上灾害管理通信系统建设领先的国家,相关经验对我国灾害管理通信系统的建设有着重要的借鉴意义。都道府县防灾无线网是日本灾害管理通信系统的重要组成部分,已具备了较为成熟的建设模式和应用经验。本文介绍了日本灾害管理通信系统的框架、都道府县防灾无线网的体系架构及其通信系统组成,选择了神奈川县、长野县和三重县作为典型案例,对这三个县的防灾无线网的建设和应用情况作了分析,并借鉴日本的经验,结合我国的实际,提出了五个方面的建设建议。     

高校档案资源收集策略

高校档案资源有功能多样性、内容学术性、收集时限性等特点。针对这些特点,文章阐述了大学档案馆所采取的收集策略,同时指出健全档案资源收集网络制度,确保高校档案及时、完整归档。     

关于大都市灾害管理之研究——以日本大都市灾害管理为中心

本文以日本政府及东京等大城市灾害管理政策、行政体制及实例为对象,分析研究了大城市在自然灾害管理中所应注意的问题和经验教训,为大城市的灾害管理提供借鉴.     

电子文件的科学管理

社会信息化是当代社会的一种大趋势 ,它正推动着公共行政现代化的进程。档案馆自动化和计算机设计技术的普及对电子文件的收集、鉴定、保管、开发利用等方面起到了一定的冲击 ,我们必须在基于传统的档案管理方法 ,以变通和发展的观点来适应新形势下的档案工作 ,更加有效的管理电子文件。     

筛选叶面喷施剂减轻旱涝灾害

本文提出使用叶面喷施剂的新概念,讨论了其作用和发展前景。观察了国内外10种药物对减轻旱涝灾害的效果,其中2种药物同田试验结果较日本名牌产品爱多收效果略好,2种药物较台湾健生素效果略好。2种属国内最新研制的新药。在与国内同类产品稀土、增产菌,903、多效唑、802、喷施宝、麻果素的同田对比中领先。     

我国建筑消防现状

近年来 ,中国城市建筑火灾发生率呈上升趋势 ,仅 1999年 ,全国共发生建筑火灾 18万起 ,比前一年增长 4 1% ;死亡 2 70 0多人 ,比前一年增长 15 % ,直接经济损失多达 14亿元。     

Earthquake Risk Perception in Bucharest, Romania

The Municipality of Bucharest is one of the capitals with the highest seismic risk in the world. Bucharest is particularly vulnerable to seismic hazard due to: the high density of inhabitants, especially within the residential districts with blocks of flats; the old public utility fund; the out-of-date infrastructure; the numerous industrial parks that are undergoing a restructuring process, not to mention the inefficient organization of civil protection and poor education of the population regarding seismic risk. This research was designed to examine the attitudes and perceptions of people living with the risk of an earthquake hazard in Bucharest. We were interested in how attitudes and perceptions differ depending on gender, age, education, residential area and socioeconomic status, characteristics of seismic hazard, degree of risk exposure, degree of danger, and casualty awareness. At the same time, we compare the results of this study with those from a previous and similar enquiry in 1997. The statistical processing has indicated a significant difference between the declared perception of seismic risk and the independent variables of gender, age, level of education, level of attachment to the residential area, and degree to which the subjects consider they may be affected and could retrieve their losses. Due to the continuous decrease of their living standard, the most vulnerable is the aged population. The feelings toward the residential area is another factor of statistical significance for the population's seismic danger perception. A strong affective bond offers a feeling of safety and leads to the neglect and even total denial of the hazard. In the case of independent variables regarding the type of dwelling, its age, and property form, deviations of empiric values from the theoretical distribution are not relevant for the correlation searched for, which indicates that this issue goes beyond the above-mentioned criteria.